Calorimetric studies in the halide-hydrogen chloride systems

Calorimetric studies in the halide-hydrogen chloride systems. R. L. Benoit, M. Rinfret, and R. Domain. Inorg. Chem. , 1972, 11 (11), pp 2603–2605. D...
0 downloads 0 Views 406KB Size
Inorganic Chemistry, Vol. 11,No. 11, 1972 2603

HALIDE-HYDROGEN CHLORIDE SYSTEMS the Kaa peak, and another near the K3p peak. It would be very desirable to have further runs on other salts of the nitrate ion. In summary, there is a moderate correlation between the calculated MO levels and XPS peaks for the nitrate ion. This was also found to be true in the case of the cyanide ion. In both anions the CNDO(1) method gives the best overall agreement with the experimental XPS energies. However, this could very well be fortuitous. I n the case of the cyanide MO calculations, the CNDO calculations come the closest

to the ab initio calculation, with the INDO calculation seemingly having trouble with the lowest binding energy levels. Both of the cyanide extended Huckel treatments give poor agreement with the ab initio calculation. The same qualitative trends are also evident in the nitrate ion MO calculations. Acknowledgments.-This work was supported in part by National Institutes of Health Grant HEWPHS-HE 13652; AEC support of the Berkeley X P S facilities is gratefully acknowledged.

CONTRIBUTION FROM THE D~PARTEMENT DE CHIMIE, UNIVERSIT~ DE MONTR~AL, MONTR~A 101, L QUEBEC,CANADA

Calorimetric Studies in the Halide-Hydrogen Chloride Systems BY R. L. BENOIT,* M. RINFRET,

AND

R. DOMAIN

Received March 29, 1972 The free energy and enthalpy of the reactions X-(solv) -I- HCl(so1v) = HClX-(solv) were determined a t 30" from calorimetric measurements in sulfolane, a very weakly basic dipolar aprotic solvent. X-(solv) denotes solvated chloride, bromide, and iodide ions. From the heat of solution of gaseous hydrogen chloride, the following enthalpy changes are obtained for the reactions X-(solv) HCl(g) = HClX-(solv): 13.9, 10.7, and 8 kcal mol-', respectively, for C1-, Br-, and I-. These enthalpies which are minimum values for hydrogen bond energies for HClX- are compared with values previously predicted from reported measurements on some solid halide salts-hydrogen chloride systems.

+

Introduction Hydrogen dihalide HX2- anions' are well characterized in salts whose lattice is stabilized by large cations such as tetraalkylammonium. The mixed HXY- anions have received less attention. Attempts have been made to evaluate the corresponding hydrogen bond energies. I n aqueous solutions, HF2- is wellknown but there is no evidence for the other ions which are apparently too unstable. However, in some nonaqueous solvents, the equilibria X-(solv)

+ HY(so1v) = HYX-(solv)

(1)

have been studied for HCL- and H B ~ Z - .Dipolar ~~~ aprotic solvents, in particular, favor the formation of hydrogen dihalide ions because the differences in solvation energies for small (X-) and large (HYX-) anions4 are considerably smaller in these solvents than in protic solvents Among the dipolar aprotic solvents, sulfolane was selected for a study of reactions 1 (X = C1, Br, I ; Y = Cl) as i t behaves as a very weak base5 so that HCl does not appreciably dissociate, its dissociation constant is estimated to be ~ 1 O - l ~ Results . of a preliminary nmr study were recently published.6 The present calorimetric investigation was carried out to provide thermodynamic data for the reactions involving the formation of HClZ-, HClBr-, and HClI- ions. Such I

(1) D.G.Tuck, P ~ o g rInorg. . Chem., S, 161 (1968). (2) I. M.Kolthoff, S. Bruckenstein, and M. K. Chantooni, Jr., J . Anzer. Chem. Soc., 88, 3927 (1961). (3) H. F. Herbrandson, R . T. Dickenson, Jr., and J. Weinstein, ibid., 1 6 , 4046 (1954). (4) G. Choux and R . L. Benoit, i b i d . , 81, 6221 (1969). (6) R . L. Benoit, C. Buisson, and G. Choux, Can. J . Chem., 48, 2353 (1970). (6) R. L.Renoit, A. L. Beauchamp, and R . Domain, Inorg. Nucl. Chem. Lett., 7,557 (1971).

data are combined with the heat of solution of HCI in sulfolane to lead to values which are related to hydrogen bond energies and can be compared with previous estimates. Experimental Section Materials.-Sulfolane was purified according to a procedure described p r e v i ~ u s l y . ~Tetraethylammonium chloride and bromide and tetrabutylammonium iodide were Eastman Kodak White Label products. Both compounds were recrystallized from ethanol and dried under vacuum a t 40". Solutions of the halide salts were prepared by weighing. Concentrated solutions of hydrogen chloride were obtained by saturation with Matheson anhydrous HCl gas after passage through a Dry Ice-acetone trap. The HCl concentration was determined by titration with standardized NaOH after flooding of the sample with water. Apparatus and Procedure.-Heats of reaction were measured with an LKB-Produkter AB (Stockholm, Sweden) Model 8725-2 Isoperibol calorimeter. It consists of a 25-ml Pyrex reaction vessel equipped with a 1700 ohm thermistor, a 50 ohm heater, and a sample-holder stirrer. Sealed 1-ml glass ampoules filled with a 3 M HC1 solution were attached to the holder. The reaction vessel containing the halide solution is enclosed in a chromeplated submarine jacket, which, in turn, is kept in a thermoregulated bath maintained a t 30.016 ?C O.O0lo. The calorimeter is calibrated electrically before and after each run by means of a Hewlett-Packard 740B standard calibrator. The current pulse into the heater is predetermined by a H P 5214L preset electronic timer followed by a flip-flop circuit driving a high-speed miniature Clare mercury wetted relay with a response time less than 1 msec. The thermistor temperature sensor is connected in series with a Guildline 9801T temperature controlled four-terminal resistor (calibrated t o 1 ppm) and a Keithley 225 constant-current source (set a t 200 PA). The voltage drops a t the thermistor and the four-terminal resistor are compared with a Guildline 9810L isolating potential comparator The difference signal between the thermistor and the resistance settings, in increments of 10 ohms (0.125'K), is recorded on a strip chart recorder, yielding a complete time-resistance curve.

.'

(7) T. M. DauphinBe, Can. J . Phys., 81,577 (1953).

2604 Inorganic Chemistry, Vol. 11, No. 11, 1972 Heat calculations were made according t o Dickinson's method. The extrapolations were carried out to the time corresponding t o half the temperature rise in the case of electrical calibration measurements and to the time corresponding t o 0.6 the temperature change in the case of heat of solution measurements. The performance of the calorimeter was checked by measuring the integral heat of solution of KC1 in water and the heat of neutralization of tris(hydroxymethy1)aminomethane with aqueous 0.1 M HC1. In both cases, results fell within less than 0.3y0 from reported v a l ~ e s . ~ ~ 9 Solution vapor pressures were measured with a precision quartz gauge (Texas Instruments) whose quartz spiral was thermostated a t 60". The gauge was connected to a 70-ml cell by a thermally jacketed glass capillary. The cell, coitaining the magnetically agitated HC1 solution, was kept in a water bath regulated to f0.005'.

Results The heat changes measured relate to the reaction between a known weight, corresponding to less than 1 ml, of 3.0 M HC1 solution and a known weight, equivalent to approximately 26 ml, of halide solutions. The concentrations CX' of the latter solutions were respectively 0.088 M for C1-, 0.096 M for Br-, and 0.089 M for I-. The calorimetric measurements were thus made a t nearly constant ionic strength. The water concentration of the solutions, as determined by the Karl Fischer method, was 8 X M . The solution volume V after reaction is calculated from the weight and density measurements. The concentration CH of acid was determined after reaction by titration. The calorimetric results are presented in Table I TABLE I CALORIMETRIC DATAFOR THE SYSTEMS HCl Vol/ml

Mmol HC1

+ X-

Mmol X -

Q/cal

BENOIT,RINFRET,AND DOMAIN

=

(HClX-) (HCl)(X-)

Q = (HClX-) X V X AH

=

f(K,CH,Cx)V X AH

The measured heat changes are corrected for the small heat of dilution of 3 M HCI estimated as -0.05 kcal mol-'. Approximate values of AH are obtained from the results corresponding to low concentrations of acid ; they in turn lead to approximate values of K . Further calculations were carried out using a computer program similar to one previously described. lo The approximate values of AH and K are refined simultaneously by the nonlinear least-squares method in order to minimize the residual function

c[(F

-

oalod 2

V

Q-)]

A few least-squares cycles resulted in convergence and the refined values of K and AH are given in Table 11. TABLE I1 THERMODYNAMIC RESVLTSFOR THE EQUILIBRIA X-(SOLV) HC~(SOLV) HC~X-(SOLV)

+

HClX-

- A H , kcal mol-'

HC128.09 j=0.18 HClBr4.93 + 0.24 HClI2 . 3 =t0 . 8 a See reference 6.

log K (cal)

log K ( R I V I N ) ~

2.54 =k 0.15 2 . 3 =I= 0 . 2 1 . 2 =I= 0 . 4

2.4 i 0 . 3 2 . 1 3z 0 . 3 0.9 A 0.3

25,54 26.09 25,98 26.04 26.22 25.88 26.47 26.75 26.60

(a) X 0.460 0.722 0.860 1.333 1.596 1 922 2,482 3,133 3.211

C1 2.253 2.340 2.331 2,271 2.277 2,244 2.266 2.337 2.318

3.70 5.61 6.69 10.26 11.80 13.82 15.90 17.55 17.34

The errors set for the 99% confidence limit are equal to three times the standard deviation estimated from the diagonal elements of the inverse matrix corresponding to the last least-squares cycle. The low accuracy obtained for HClI- is due to the low heat changes measured. The results of the previous nmr study6are also quoted in Table 11. The data of the vapor pressure measurements are given in Table 111. The sulfolane vapor pressure is

26.28 26.23 26.54 26.64 26.75 27.08 28.00 27,32

(b) X = Br 0.497 2.507 0.754 2.501 0.944 2.521 1.421 2.514 1,658 2.514 2.398 2.517 2.479 2.515 3.253 2.517

2.30 3.51 4.24 6.45 7.41 9.57 9.86 10.91

TABLE I11 VAPORPRESSURES OF HCl SOLUTIONS /Temp, "C 29.53 36.44 40.3 0.83MHC1 \ p , m m 172.3 214.7 240.9 (Temp, "C 27.42 30.95 35.84 41.26 1 , 7 6 M HCl 353.0 392.6 486.0 536.9 :$:e) "C 26.76 32.25 37.87 41.58 2.29 M HC1 \p, mm 464.7 554.1 662.1 742.1

25.72 25.65 25.76 26.10 26.35 26.20

(c) 0.522 0.853 1,423 1.966 3.015 3.445

0.86 1.16 1.66 2.08 2.82 3.08

less than 0.05 mm a t the temperatures considered so that the HC1 partial pressure is taken as the total vapor pressure. From these results the heat of solution of HCI Corresponding to HCl(g) = HCl(so1v) is calculated as AH = -5.8 f 0.1 kcalmol-l.

x

=

=

I 2.256 2.260 2.260 2,263 2.261 2,252

which includes solution data and corresponding heat changes Q. The equilibrium constants K and enthalpies AH of reaction 1 are related to concentrations by the following equations. (8) J, Greyson and H. Snell, J. P h y s . Chem., 73,3208(1969). (9) J. 0.Hill, G. 6jelund. and I. Wadsb, J. Chem. Thermodyn., 1, 111 (1969).

Discussion The values obtained here for K , the formation constants of the HClX- ions, are in good agreement with the values determined in the nmr study under somewhat different experimental conditions.6 It can only be inferred that K is not very sensitive to the presence of (10) R. L. Benoit, A. L. Beauchamp, and M. Deneux, 73, 3268 (1969).

J. P h y s . Chem.,

HALIDE-HYDROGEN CHLORIDE SYSTEMS

Inorganic Chemistry, Vol. 11, No. 11, 1972 2605

and Vall4elZand Deiters13 and are listed in the second small amounts of water (2 X lov2 t o 9 X M ) and column of Table IV. These authors determined the to variations in ionic strength (0.09 to 0.6 M ) . enthalpy changes for reactions Comparisons with literature values for K H Cin~other ~ solvents have already been made.6 The value of the R ~ N X ( S+ ) HCl(g) = R4NHClX(s) (4) formation constant of HC12- can be used to estimate by calorimetric and manometric methods. It was the stability of HC12- in water. Combining (HC12-),/ argued that the lattice enthalpies of R4NX and R4((CI-)s(HCI)s) = 102.6with P H C l = k(HCI),, where k = NHClX would tend to be equal a t some large cation 0.277 when p is expressed in atmospheres and (HCl), is in moles per liter, gives (HC12-)s/(C1-)s = 1 0 3 . z # ~ ~ 1 .size so that under these conditions, the enthalpy change of reaction 4 would approach the energy of the hydrogen If we now assume that the energy of transfer of the bond in HClX-. As expected, i t was found that the related ions HCl2- and AgClz-, from water to sulfolane, heats of reaction of HC1 with R4NC1 salts increased in are not too different, the previously reported yalues'O the order Me4N (11.3-12.0 kcal mol-'), Et4N (13.7 (AgCl~-)s/(Cl-)s = lo'.' and (AgC12-)w/(Cl-)w = kcal mol-'), Bu4N (12.6-14.2 kcal mol-'). This latter 10-4.4 lead to (HC12-)w/(Cl-)w = 10-2~9#Hc1.Thus "limit value" 14.2 kcal mol-' was thus taken as a a t 30°, in a 10.8 M aqueous HC1 solution (#HC1 = minimum for the hydrogen bond strength in HCL-. 0.10), the concentration of HCL- ions would be The values for HCIBr- and HClI- given in Table IV 10-2.gM . Although such values are only approximate, were also obtained for the tetrabutylammonium salts. they a t least explain qualitatively why HC12- has not The approximate agreement between both sets of been detected by Raman spectroscopy in aqueous soluAH values in Table I V is coincidental as the discrepantions even in the presence of added chloride." cies merely reflect the apparently small differences (less No AH values for the solvated HClX- ions appear than 2 kcal mol-') in heats of solution of the simple t o have been reported to date.' More fundamental and complex halide salts. The validity of the assumpenthalpy values can be deduced from the enthalpy tion of McDaniel and Vall4e,l2 i.e., that the lattice changes for reaction 1 by adding the heat of solution enthalpies of Bu4NC1 and Bu4NHCI2 are nearly equal, of gaseous HCl in sulfolane, HCl(g) = HCI(so1v) is however open to question. Although the ionic radii 5.8 kcal mol-'. The calculated enthalpies which corof R4N+ and HC12- are in doubt, the simple Kapustinrespond to the reactions skii formula indicates some appreciable difference in X-(solv) HCl(g) = HClX-(solv) ( 2) lattice enthalpies. Furthermore, later studiesI4 with a similar assumption led to a value of -37 kcal mol-' are given in the first column of Table IV for the hydrogen bond strength in HF2- while Waddington's extensive calc~lations'~ gave -58 + 5 kcal TABLE IV mol-'. I n summary, i t can only be said that both sets MINIMUM HYDROGEN BONDENERGIES HClX---AH, kcal mol-l--of figures given in Table IV represent minimum values HCL13.9" 14.2b for the hydrogen bond strength in HClX- anions, the HClBr10.7" 9.2" real values may be considerably higher but remain unHClI8" 8.& attainable as long as solvation enthalpies of X - and a This work. Reference 12. Reference 13. HCIX- ions or lattice enthalpies of R4NX and R4NHClX cannot be predicted. Although direct gasThe enthalpy changes for reaction 2 are closely rephase determinations of hydrogen bond strength in lated to the hydrogen bond energies defined for HClX- have recently been attempted,16 they do not appear t o be free from difficulties; the following values X-(g) + HCl(g) = HClX-(g) (3 ) were obtained by extrapolation: HClz-, 24; HClBr-, Although these latter values cannot be calculated as 16 kcal mol-'; while 30 G= 5 kcal mol-' was recomthe solvation enthalpies for the nonspherical HClXmended for HFz-. The first two values are respecgaseous ions are not known, i t is expected, on account tively 10 and 5 kcal mol-' higher than those obtained of ion sizes4 alone, that the solvation enthalpies of for equation 2. These differences, which represent HClX- are smaller than that of X-, so that the AH the differences in solvation enthalpies between C1values for reactions 2 represent minimum values for and HCLand C1and HClBr-, are not unreasonable the hydrogen bond energies corresponding to reactions when the solvation enthalpies already quoted in the 3. Some estimqtes of solvation enthalpies of halide text for C1-, Br-, I-, and Clod- are considered. and perchlorate ions,4 respectively (Cl-, 80.1 ; Br-,

-

+

77.0; I-, 72.1; C104-, 62 kcal mol-'), indicate that the hydrogen bond energies may be substantially higher than the very minimum values quoted in Table IV. Previous estimates of hydrogen bond energies for HClz-, HCIBr-, and HClI- were made by McDaniel (11) A. G.Maki and R. West, Inovg. Chem., 2 , 667 (1963).

Acknowledgments.-We thank the National Research Council of Canada for financial support and Dr. A. L. Beauchamp for the use of the computing program. (12) D H.Mcnaniel and R. E. Vallee, zbrd., 2, 996 (1963). (13) R.M. Deiters, Ph.D. Thesis, as quoted in ref 1. McDaniel, I. J . Amev. Chem. Soc., 86, 4497 (14) S . A. Harrell and D. € (1964). (15) T. C. Waddington, Trans. Favaday Soc., 2 6 , 64 (1958). (16) R.Yamdagni and P.Kebarle, J . Amev. Chem. Soc ,98,7139(1971).